Development of State Duration and Expectation Model for Evaluating Remaining Life of Manufacturing Systems

Abstract

The key goal of predicting the state expectation of an engineering system is to predict the remaining life of the system so as to aid in maintenance decision-making activities. In this paper, a maintenance model for predicting the state expectation of industrial machines has been developed. It incorporates various stages of deterioration and maintenance states. Given that a current state has been attained, from inspection and diagnosis, this model is capable of computing the predicted average time before a system failure occurs. This study focuses on using real data of an industrial Bottle filler machine to test the effectiveness of the State Expectation model and its effect on the reliability and maintainability of the machine. The model is tested for various scenarios by changing one of the main parameters during each calculation while others are kept constant. For the state duration sensitivity analysis, as the failure rate continuously increases from 0.0178 to 0.7060, the expected mean sojourning time for each degradation state decreased from 220.97hrs to 1.59hrs. Subsequently at uniform incrementally varied repair/maintenance rate (0.0861 to 0.7663), the state expectation of the equipment increases from 168.11hrs to1189.98hrs. This allowed us to determine the most suitable decision to improve the reliability of the Bottle filler machine. The prediction result identifies the effectiveness of the proposed method in predicting RUL of manufacturing systems.

Keywords: Remaining Useful Life; State Expectation Model; Maintenance; Decision making; Prediction.